Lipschitz distance between clouds
Ivan N. Mikhailov · 2025
Read Lipschitz distance between clouds by Ivan N. Mikhailov (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-22696.txt.
Introduction
Ivan N Mikhailov, writing alone in a 2025 mathematics note, shows that the Lipschitz distance between two classes of metric spaces is positive. One class lies at finite Gromov-Hausdorff distance from a one point space, and the other from the real line with its natural metric. The positive result means the two classes stay genuinely apart under this stronger stretch based comparison.
Catalog record
| CREATOR | Ivan N. Mikhailov |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-22696
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Lipschitz distance between clouds Ivan N. Mikhailov Abstract In this note we show that the Lipschitz distance between the classes of metric spaces at finite Gromov–Hausdorff distances from the one-point metric space Δ 1 \Delta_{1} and the real line ℝ \mathbb{R} with the natural metric, respectively, is positive. 1 Introduction In [ 3 ] M. Gromov introduced moduli spaces of the class of all metric spaces at finite Gromov–Hausdorff distances from a given metric space. It was mentioned that such moduli spaces are always complete and contractible ( [ 3 ] [section 3.11_{\frac{1}{2}_{+}} ]). In [ 1 ] the authors suggested to work with such moduli spaces (they were called clouds ) in the sense of NBG set theory to avoid arising set-theoretic issues. While the completeness of each cloud was verified in [ 1 ] , the contractibility of each cloud remains an open question for a number of reasons. The main issue here is that a natural homothety-mapping that takes a metric space (X,d_{X}) into (X,\lambda d_{X}) for some \lambda>0 and generates a contraction of a cloud of all bounded metric spaces if \lambda\to 0 , does not behave so well in case of unbounded metric spaces. Firstly, in [ 1 ] it was shown that there exist metric spaces such that d_{GH}(X,\lambda X)=\infty for some \lambda>0 . The simplest one is a geometric progression X=\{3^{n}\colon n\in\mathbb{N}\} with a natural metric, for which d_{GH}(X,2X)=\infty . Secondly, even for clouds that are invariant under multiplication on all positive numbers a homothety-mapping may not be continuous. In [ 6 ] it was shown that d_{GH}(\mathbb{Z}^{n},\lambda\mathbb{Z}^{n})\geq\frac{1}{2} for all \lambda>1 , n\in\mathbb{N} . In [ 7 ] the global geometry of clouds was studied. Namely, each cloud equipped with the Gromov–Hausdorff distance is a metric class, and, hence, the Gromov–Hausdorff distance between different clouds can be defined. It was shown [ 7 ] [Corollary 3 ] that the Gromov–Hausdorff distance between the cloud of bounded metric spaces and the cloud of the real line is infinite. In this note we consider the Lipschitz distance between clouds instead of the Gromov–Hausdorff distance. By the idea of Alexey Tuzhilin, since the homothety-mapping is continuous with respect to the Lipschitz distance, this trick could allow to develop a sound geometry of clouds. However, a tricky problem appears immediately: is there a pair of clouds on a finite and non-zero Lipschitz distance from each other? By applying the ideas from [ 7 ] , we show that the Lipschitz distance between the cloud of bounded metric spaces and the cloud of the real line is positive. However, the question whether this distance is finite remains open. 2 Preliminaries A metric space is an arbitrary pair (X,\,d_{X}) , where X is an arbitrary set, d_{X}\colon X\times X\to[0,\,\infty) is some metric on it, that is, a nonnegative symmetric, positively definite function that satisfies the triangle inequality. For convenience, if it is clear in which metric space we are working, we denote the distance between points x and y by |xy| . Suppose X is a metric space. By U_{r}(a)=\{x\in X\colon|ax|<r\} , B_{r}(a)=\{x\in X\colon|ax|\leq r\} we denote open and closed balls centered at the point a of the radius r in X . For an arbitrary subset A\subset X of a metric space X , let U_{r}(A)=\cup_{a\in A}U_{r}(a) be the open r -neighborhood of A . For non-empty subsets A\subset X , B\subset X we put \operatorname{d}(A,\,B)=\inf\bigl\{|ab|:\,a\in A,\,b\in B\bigl\} . 2.1 Hausdorff and Gromov–Hausdorff distances Definition 1 . Let A and B be non-empty subsets of a metric space. The Hausdorff distance between A and B is the quantity d_{H}(A,\,B)=\inf\bigl\{r>0\colon A\subset U_{r}(B),\,B\subset U_{r}(A)\bigr\}. Definition 2 . Let X and Y be metric spaces. The triple (X^{\prime},Y^{\prime},Z) , consisting of a metric space Z and its two subsets X^{\prime} and Y^{\prime} , isometric to X and Y , respectively, is called a realization of the pair (X,Y) . Definition 3 . The Gromov-Hausdorff distance d_{GH}(X,Y) between X and Y is the exact lower bound of the numbers r\geq 0 for which there exists a realization (X^{\prime},Y^{\prime},Z) of the pair (X,Y) such that d_{H}(X^{\prime},\,Y^{\prime})\leq r . Now let X,\,Y be non-empty sets. Definition 4 . Each \sigma\subset X\times Y is called a relation between X and Y . By \mathcal{P}_{0}(X,\,Y) we denote the set of all non-empty relations between X and Y . We put \pi_{X}\colon X\times Y\rightarrow X,\;\pi_{X}(x,\,y)=x, \pi_{Y}\colon X\times Y\rightarrow Y,\;\pi_{Y}(x,\,y)=y. Definition 5 . A relation R\subset X\times Y is called a correspondence , if their restrictions \pi_{X}|_{R} and \pi_{Y}|_{R} are surjective. Let \mathcal{R}(X,\,Y) be the set of all correspondences between X and Y . Definition 6 . Let X,\,Y be metric spaces, \sigma\in\mathcal{P}_{0}(X,\,Y) . The distortion of \sigma is the quantity \operatorname{dis}\sigma=\sup\Bigl\{\bigl||xx^{\prime}|-|yy^{\prime}|\bigr|\colon(x,\,y),\,(x^{\prime},\,y^{\prime})\in\sigma\Bigr\}. Proposition 1 ( [ 2 ] ) . For arbitrary metric spaces X and Y , the following equality holds 2d_{GH}(X,\,Y)=\inf\bigl\{\operatorname{dis}\,R\colon R\in\mathcal{R}(X,\,Y)\bigr\}. 2.2 Lipschitz distance Definition 7 . A homemorphism f of metric spaces is called bilipschitz iff both maps f and f^{-1} are Lipschitz. Definition 8 . The Lipschitz distance between metric spaces X and Y is the value d_{L}(X,\,Y)=\inf_{f\colon X\to Y}\,\ln\bigl(\max\bigl\{\operatorname{dil}(f),\,\operatorname{dil}(f^{-1})\bigr\}\bigr), where infinum is taken over all bilipschitz homeomorphisms f\colon X\to Y . 2.3 Clouds By \mathcal{VGH} we denote the class of all non-empty metric spaces, equipped with the Gromov–Hausdorff distance. Note that \mathcal{VGH} is a proper class in the sense of the NBG set theory. In this theory all objects are classes of one of the two following types: sets , or proper classes . A class is called a set if it belongs to some other class, and a proper class otherwise. It is important for us that for all classes the following natural constructions are defined: Cartesian product, maps between classes, metrics, pseudometrics, etc. Theorem 1 ( [ 2 ] ) . The Gromov–Hausdorff distance is a generalised pseudometric on \mathcal{VGH} , vanishing on pairs of isometric metric spaces. Namely, the Gromov–Hausdorff distance is symmetric, satisfies a triangle inequality, though can vanish or be infinite between some pairs of non-isometric metric spaces. A class \mathcal{GH}_{0} is obtained from \mathcal{VGH} by factorization over zero distances, i.e., over an equivalence relation X\sim_{0}Y , iff d_{GH}(X,\,Y)=0 . Definition 9 . Consider an equivalence relation \sim_{1} on \mathcal{GH}_{0} : X\sim_{1}Y , iff d_{GH}(X,\,Y)<\infty . We call the corresponding equivalence classes clouds . For an arbitrary metric space X , we denote a cloud containing X by [X] . Let \Delta_{1} be a metric space, consisting of a single point. Hence, [\Delta_{1}] is the cloud of all bounded metric spaces. Suppose that for some metric spaces A and A^{\prime} , the equality holds d_{GH}(A,A^{\prime})=0 . Then, for arbitrary metric space B , we also have d_{GH}(A,B)=d_{GH}(A^{\prime},B) . From this simple observation, it follows that all the results about d_{GH}(A,B) also hold if we exchange A by A^{\prime} such that d_{GH}(A,A^{\prime})=0 . Thus, instead of interpreting << A\in[X] >> directly by definition so that A is an equivalence class of all metric spaces on zero Gromov–Hausdorff distances from each other, we will mean that A is a certain member of this equivalence class. For example, X\in[\Delta_{1}] can be read as << X is a bounded metric space>> throughout the paper. Theorem 2 ( [ 2 ] )